Abstract
Peptides that inhibit MDM2 and attenuate MDM2-p53 interactions, thus activating p53, are currently being pursued as anticancer drug leads for tumors harboring wild type p53. The thermodynamic determinants of peptide-MDM2 interactions have been extensively studied. However, a detailed understanding of the dynamics that underlie these interactions is largely missing. In this study, we explore the kinetics of the binding of a set of peptides using Brownian dynamics simulations. We systematically investigate the effect of peptide C-terminal substitutions (Ser, Ala, Asn, Pro) of a Q16ETFSDLWKLLP27 p53-based peptide and a M1PRFMDYWEGLN(12) 12/1 phage-derived peptide on their interaction dynamics with MDM2. The substitutions modulate peptide residence times around the MDM2 protein. In particular, the highest affinity peptide, Q16ETFSDLWKLLS27, has the longest residence time (t similar to 25 mu s) around MDM2, suggesting its potentially important contribution to binding affinity. The binding of the p53-based peptides appears to be kinetically driven while that of the phage-derived series appears to be thermodynamically driven. The phage-derived peptides were found to adopt distinctly different modes of interaction with the MDM2 protein compared to their p53-based counterparts. The p53-based peptides approach the N-terminal region of the MDM2 protein with the peptide C-terminal end oriented toward the protein, while the M1PRFMDYWEGLN(12)-based peptides adopt the reverse orientation. To probe the determinants of this switch in orientation, a designed mutant of the phage-derived peptide, R3E (M1PEFMDYWEGLN(12)), was simulated and found to adopt the orientation adopted by the p53-based peptides and also to result in almost a 5-fold increase in the peptide residence time (similar to 120 mu s) relative to the p53-based peptides. On this basis, we suggest that the R3E mutant phage-derived peptide has a higher affinity for MDM2 than the p53-based peptides and would therefore, competitively inhibit MDM2-p53. The study, therefore, provides a novel computational framework for kinetics-based lead optimization for anticancer drug development strategies.